Human T cell receptor complexed with class I MHC and antigenic peptide: TCR/Tax/HLA-A2 In February 1996 we collected diffraction data from two frozen native crystals (~90 K) of the TCR/Tax peptide/HLA-A2 ternary complex (beamline A1; ADSC CCD) . Because of detector hardware and software problems, neither data set was complete but merged to yield a data set (2.8 _, Rmerge = 11.6 %). Both crystals were highly mosaic (1.4-1.8x) ( space group C2, a = 228.2 _, b = 49.7 _, c = 95.7 _, b = 91.3x). A partial data set (3.3 _, Rmerge=12.1%) was collected from a crystal containing a variant peptide with cysteine at P1 soaked, in ethyl mercury phosphate. The position of the mercury atom was located with a difference Patterson. A full data set for a platinum derivative (3.5 _, Rmerge=12.3%) was also collected. Two sites were located using a difference Fourier phased on a partial molecular replacement solution . (Phasing from the platinum and partial mercury data is poor: fh/e = 0.7 Hg, and 0.9 Pt to 3.5 _, and fom= 0.3). In May 1996 two new native and five different Hg (produced by adding 1 or 2 cys to one chain of the MHC molecule) data sets were collected at A1. One native and one Hg (peptide P1 and P11-cys) crystal had the same cell dimensions as the C2 crystal but were not centered (P2), which should allow non-crystallographic and multiple crystal averaging. Human natural killer cell receptor. In May 1996, we collected a full data set from a thin (70mm) needle crystal of the ectodomain of the NK receptor. (a=81._, c=71._; P61 or P65; diffraction beyond 2.6_).